deepspeedai/DeepSpeed

Pull Requests Commits

Merge branch 'master' into s3-pr

tjruwase committed 4 years ago

Verified 6dfe6938

fix partition activations issue when mp=2 and pp=2 (#1589)

eltonzheng committed 4 years ago

Verified 51d42ab9

Merge branch 'master' into s3-pr

tjruwase committed 4 years ago

Verified 80b53d31

Fix return code on error (#1540)

Dipet committed 4 years ago

Verified 499800ca

Fix merge issues

tjruwase committed 4 years ago

2b5f6ea2

Merge branch 'master' into s3-pr

tjruwase committed 4 years ago

Verified 11d590a0

Enable AVX256 on AMD CPU (#1360)

fumihwh committed 4 years ago

Verified a637cc2c

Removing `ImportError` from tutel import try/except (#1583)

alexandremuzio committed 4 years ago

Verified 1bc13fe8

Replace brute force and add log (#1560)

chunyang-wen committed 4 years ago

Verified e2b39ded

Add documentation for TensorBoard logging (#1577)

manuelciosici committed 4 years ago

Verified e1b4aa8f

remove debug prints (#1585)

stas00 committed 4 years ago

Verified bcf2bdde

jeffra committed 4 years ago

8220674d

Several fixes for our read-the-docs build (#1579)

jeffra committed 4 years ago

Verified a8a17f23

Enables ZeRO-3 inference (#1514)

jeffra committed 4 years ago

Verified 2332cb31

align new line with master

jeffra committed 4 years ago

16f1d21d

convert to same new-line style as master

jeffra committed 4 years ago

2a383020

[CI] transformers@master has been fixed (#1573)

stas00 committed 4 years ago

Verified 74baf5bb

jeffra committed 4 years ago

6dc21a60

switch bin files to use python3 instead of python (#1185)

jeffra committed 4 years ago

Verified 236890d6

Render docs for pipe.ProcessTopology (#1505)

James Reed committed 4 years ago

Verified fafc827d

Remove hard tensorboardX requirement (#1571)

jeffra committed 4 years ago

Verified a90497ec

[launcher/runner] respect CUDA_VISIBLE_DEVICES for a single node (#960)

stas00 committed 4 years ago

Verified e3c2d7b1

[autotuning] guard tabulate package import (#1569)

jeffra committed 4 years ago

Verified 938449e3

Enforce nccl/rccl alignment of start location of each shard (#1564)

amathews-amd committed 4 years ago

Verified 4a0b1032

bump DSE commit

jeffra committed 4 years ago

4625add6

set hf hash (#1568)

jeffra committed 4 years ago

Verified da7bff40

Fix partial recovery of sparse_tensor_module_names and dynamically check if gradient data is sparse (#1562)

Dipet committed 4 years ago

Verified 4bf4ab7a

Add autotuning news post (#1565)

cli99 committed 4 years ago

Verified bda3d0e6

[build] support cuda-11.5 (#1558)

stas00 committed 4 years ago

Verified fa8d6c0a

Update offload parameter names (#1536)

tjruwase committed 4 years ago

Verified 7567c76c

Older